Posts: 13 Threads: 1 Likes Received: 9 in 7 posts Likes Given: 2 Joined: May 2020 Reputation: 0 04-May-2020, 11:42 PM (This post was last modified: 12-Jun-2020, 03:43 PM by Sir Brokalot FPV.) Guys i need help tuning this quad. its driving me nuts.. I keep getting oscillations everywhere and i'm resorting to blackbox to try to find out where the problem is. Things i've tried: - Change props - Physically check motors for damage or loose bearings (2 rear motor bells seem to be movable slightly, but not much) - Use betaflight motor tab to determine if a motor is causing vibrations (all beneath 0.03) - Replaced motor nuts because some where off balanced due to many crashes - Applied Joshua's RPM filter setup - Tried increasing/decreasing P and D Thus far I havent been able to figure out what is causing these vibrations. I hope some of you can tell me more by looking at my blackbox log: https://www.dropbox.com/s/lg4dxszvee9nej...g.BFL?dl=0 Please advise... My gear: LEACO UMBRELLA 5 Inch Holybro Kakute F7 v1 + Tekko32 4 in1 35A ESC's Racerstar SPROG X 2206 Here is a screenshot of the part where oscillations are really visible in the video footage: Here a video of the flight: • Posts: 4,731 Threads: 392 Likes Received: 3,231 in 1,827 posts Likes Given: 3,214 Joined: Apr 2019 Reputation: 101 05-May-2020, 12:19 AM (This post was last modified: 05-May-2020, 12:21 AM by the.ronin.) Haven't looked through your traces yet cos wanted to make sure you took a look at your TPA on your Rate Profile tab. That's usually the first on scene when it comes to throttle induced oscillations. [edit] TPA or Throttle PID Attenuation does just that - attenuates your PIDs in proportion to throttle input. • Posts: 13 Threads: 1 Likes Received: 9 in 7 posts Likes Given: 2 Joined: May 2020 Reputation: 0 05-May-2020, 12:23 AM (This post was last modified: 05-May-2020, 12:27 AM by Sir Brokalot FPV.) (05-May-2020, 12:19 AM)the.ronin Wrote: Haven't looked through your traces yet cos wanted to make sure you took a look at your TPA on your Rate Profile tab. That's usually the first on scene when it comes to throttle induced oscillations. [edit] TPA or Throttle PID Attenuation does just that - attenuates your PIDs in proportion to throttle input. Thanks for the tip, I actually havent changed TPA at all, its still at standard, what would you recommend? I'm gonna do some research about it right now.. • Posts: 4,731 Threads: 392 Likes Received: 3,231 in 1,827 posts Likes Given: 3,214 Joined: Apr 2019 Reputation: 101 05-May-2020, 12:28 AM (This post was last modified: 05-May-2020, 12:30 AM by the.ronin.) Up. LOL sorry but I'm not kidding. Edge up your TPA from that .65 say by incremements of .05 and do some punchouts. I find with mid throttle oscillations is I can really only see them in GoPro video not in the FPV feed or of course in Blackbox tracing. So I'd do a whole bunch of runs at different TPAs and see where you end out. I'd caution you though maybe try and undo all those other things you did so you start out clean. Of course assuming you can remember what you did. [edit] Just saw you posted your blackbox log I'll download and take a look later I'm curious to see it. • Posts: 4,731 Threads: 392 Likes Received: 3,231 in 1,827 posts Likes Given: 3,214 Joined: Apr 2019 Reputation: 101 Also, 4S? Off the bat I can tell you your D's are generally high even for a 4S. For 6S, your motors are smokin hot? You have a lot of headroom left on your Ps. Same with your Feedforward (should be at least 100 if not 200+). General first smell check on your PIDs is that your Ds are too high and Ps too low. • Posts: 4,731 Threads: 392 Likes Received: 3,231 in 1,827 posts Likes Given: 3,214 Joined: Apr 2019 Reputation: 101 05-May-2020, 12:48 AM (This post was last modified: 05-May-2020, 12:48 AM by the.ronin.) I took a quick look at your traces. Here is what one of your punchouts looks like (see the stick overlay movement of throttle at position 1578). You can see how the D (yellow) is overpowering your P (red) causing your quad to oscillate (gyro = thick blue) relative to what your stick commands want it to do (setpoint = thick orange). In comparison, here is a trace of one of my recently tuned quads punchots (stick overlay showing throttle at 1648). • Posts: 4,731 Threads: 392 Likes Received: 3,231 in 1,827 posts Likes Given: 3,214 Joined: Apr 2019 Reputation: 101 Sorry last thing ... I actually would revisit your PIDs and filtering before you touch your TPAs. The TPA settings for the quad pictured above is default. • Posts: 13 Threads: 1 Likes Received: 9 in 7 posts Likes Given: 2 Joined: May 2020 Reputation: 0 05-May-2020, 01:09 AM (This post was last modified: 05-May-2020, 01:13 AM by Sir Brokalot FPV.) Yeah im flying a 4S. hm thanks for pointing that out. I have shifted around the P and D settings but didn't really see any improvement. Motors are warm but not hot, even when using less filtering. Ok im gonna try to raise P and set D and filtering back to default and start adjusting it from there. Thanks allot for taking a look at my log. I'm quite new with blackbox so i didnt really know what to look for.. I'll post an update after testing it. Posts: 4,731 Threads: 392 Likes Received: 3,231 in 1,827 posts Likes Given: 3,214 Joined: Apr 2019 Reputation: 101 I recommend turning off D-Min while you test. For most of the flight you will be at your D-Min. When you are comfortable and still want to use D-Min, then turn it back on, copy your D values to the D Min column (it goes away if you turn it off) and copy your P's to your D's. Don't worry since you will be using D-Min most of the time and only hit your D's on hard moves (based on your D-Min Gain). Also turn off your Feed Forward for now. Start with your pitch and roll D's at 35-ish and start bumping up your P's in increments of 5 or even 10. It can be as much as 2x or even more your D. Watch for oscillations then scale it back. Remember these Ps and Ds for your pitch and roll - it's the ratio that is important not the actual numbers. As you turn back on Feed Forward and D-Min and everything else, you can change your Ps and Ds however you want so long as you keep this ratio. For yaw or I don't worry about it too much, it's very hard to "break" those. I'm really surprised your motors weren't more hotter. • Posts: 13 Threads: 1 Likes Received: 9 in 7 posts Likes Given: 2 Joined: May 2020 Reputation: 0 05-May-2020, 01:47 AM (This post was last modified: 05-May-2020, 01:47 AM by Sir Brokalot FPV.) Thanks ill try what you said and start tuning my pids from your suggested starting values. Posts: 4,731 Threads: 392 Likes Received: 3,231 in 1,827 posts Likes Given: 3,214 Joined: Apr 2019 Reputation: 101 Just always check motor temps after each run ... something I ALWAYS forget lol. • Posts: 13 Threads: 1 Likes Received: 9 in 7 posts Likes Given: 2 Joined: May 2020 Reputation: 0 06-May-2020, 09:03 PM (This post was last modified: 06-May-2020, 09:35 PM by Sir Brokalot FPV.) ok i've been trying all kinds of adjustments but i just can't seem to find the issue. After lowering my P's and D's i still had the exact same oscillations, however it seems that the traces in blackbox look slightly cleaner for the second flight. log00003.bfl : https://www.dropbox.com/s/a6bs8lfs8fvwgz...3.BFL?dl=0 After doing some more research i've decided to lower my D Term Filter Multiplier to cover the peak frequency i measured in blackbox. D Term Filter Multiplier has been lowered and Gyro Filter Multiplier has been raised. After testing I did notice the shakes during low throttle flying were much less. However, it seems that every time my throttle reached around 1500 the oscillations return, plus i'm getting a big bounceback from lack of D it seems. (Note: i've also raised my TPA Breakpoint to 1500, could that be related?) Could it be some kind of resonance in the frame causing this, or maybe just badly balanced motors? Where to go from here? See the screenshots and PID's during mid throttle here: Video from second flight here: (audio warning sound is loud) • Posts: 4,731 Threads: 392 Likes Received: 3,231 in 1,827 posts Likes Given: 3,214 Joined: Apr 2019 Reputation: 101 Your D Min is way low dude. On 4S you should be starting out at 40s - 60s. Your P's should be 1.5-1.75x that maybe even 2x. Set your D Max to the same as your P's. Try that please - CHECK MOTOR TEMPS ALWAYS. • Posts: 13 Threads: 1 Likes Received: 9 in 7 posts Likes Given: 2 Joined: May 2020 Reputation: 0 I tried that before i changed the filtering and my quad started getting high freq oscillations with P at around 75. But ill try what u suggested tomorrow and ill update on my progress. Thanks again • Posts: 4,731 Threads: 392 Likes Received: 3,231 in 1,827 posts Likes Given: 3,214 Joined: Apr 2019 Reputation: 101 07-May-2020, 12:18 AM (This post was last modified: 07-May-2020, 12:26 AM by the.ronin.) You might want to just disable D-Min for now and run with fixed D's to remove a layer of complication out of your testing. Once you've got something you like then you can reintroduce D-Min. You are basically flying most of the time at your D-Min and only hitting D's on sharp moves and even that is dependent on your D-Min Gain. So basically in your upper left view, you are flying at P=30 D=11 which is really aggressive on P. On hard moves you move to P=30 D=20 which is now too little P. Hope that makes sense. [edit] Or just put these in for P, D, and D-Min (for 4S system) 60, 60, 40 and go from there raising your Ps. If you are getting prop wash oscillation (do a HARD 180 turn), start with P=75, D=75, D-Min=50 and increase P from there. You can move D-Min as high as 60 and multiply by 1.5 for P and D but be careful going past 60 as your starting points. And def check motor temps. Hope this makes sense. • |